Tuesday, 21 May 2013

Planck 2013 Live Blog: Day One

Okay, this post technically isn't a live blog.  I unfortunately missed the morning sessions, due to a combination of jet lag, lack of sleep and not being able to find the conference.  I did manage to attend the first session after lunch (the third plenary session), but I did not have internet access at that time as I hadn't registered.  But I did take notes, with the intention of putting them up later.  Conference website.

We had three talks: a good if slightly unclear advert for a linear collider; an inflationary talk that I found a bit hard to follow; and a talk that was half software-advert, half the-CMSSM-isn't-ruled-out-yet-honest.

2:30 pm: Gudrid Moortgat-Pick, "Do we need an LC to see SUSY?"

Higgs discovery is a revolution; not yet clear if it is the SM Higgs.  Strong limits on coloured SUSY; what physics space is left?  Do we need a Linear collider?

An LC can give us excellent resolution, including measuring top mass to 0.1 GeV and Higgs couplings and total widths to a few percent level.

Japan's leading action on this front makes it a relevant and interesting question right now.  The idea is to start with a 250 GeV Higgs factory, and evolve to a 500 GeV top factory later.  Construction could start as early as 2015.

SUSY: Constrained _and_ simplified models are under tension.  But note that CMSSM has the well-known problems, and simplified models are in some sense optimal cases.  The anomalous muon g−2 points towards low scale SUSY masses.  All(!) we need do is decoupled the coloured and uncoupled sectors in some way.

Naturalness and the need for a small μ parameter; this is well known.

I think the extent to which LHC limits can be avoided is being strongly oversold here.

Light, heavily mixed stop: an LC can (fairly) easily measure the mixing angle using a left-right asymmetry.  In particular, only with the LC can we compare the Higgs mass and stop mixing angles.

Light Higgsinos (small μ parameter) means the lightest chargino and two lightest neutralinos are essentially degenerate.  This makes it very hard to resolve the different states at the LHC (soft pions?  It is to laugh!)  But an LC can make use of ISR to measure the masses and look for different resonances.  The μ parameter can be resolved to the GeV-level in the relevant domain.

An additional source of accuracy comes from threshold scans.

Finally, what of the nightmare scenario?  Even if we find nothing else, there are still useful things an LC can do.  The precision measurements of e.g. mtop, mW and sinθ offer sensitivity to high energy scales a la LEP.  Is this enough to sell to governments?  An LC can measure the top mass ten times better than the LHC/Tevatron.

Note, LEP and SLD give different values for sinθ, with in particular the LEP value disagreeing with the MSSM.  But the errors are large enough that the world average is in agreement.  An LC is needed to probe this further.

The sound quality in this room isn't very good.  I mean, it's better than Planck last year, but still very poor.  I pity the students who learn in this classroom on a daily basis.

This talk seems to come at the question very much from a SUSY perspective.  Even the "what if we see nothing" questions seemed to assume the MSSM exists, it's just heavy.  Useful, perhaps, for a framework to work with but I got a somewhat different feeling.

Question: what of a 37 TeV Super-LHC?  Answer: why wait twenty years for measurements that we can do better, now, with an LC?

3:00 pm: Tim Jones, "Higgs thermal inflation and low energy supersymmetry"

MSSM plus a U(1)?

F-term hybrid inflation, reheating at 1015 GeV; followed by thermal inflation driven by a Higgs at about 109 GeV.  This fixes some problems of F-term inflation, such as …

Our U(1) is not dark, i.e. visible superfields charged under group in anomaly-free way.  Extra singlet S couples both to Higgs, and to F-term inflationary sector.  S gains SUSY-scale vev, solving μ problem.  Inflationary field gains large vev, leading to neutrino see-saw scale.

All fields beyond the MSSM get large masses in the desired vacuum, and can be integrated out.  Despite the S field having a large mass, it has zero vev before SUSY-breaking and so has a MSUSY-scale expectation.

SUSY-breaking also ensures we have the desired vacuum be the lower energy one, i.e. the inflationary fields get the large VEVs and the Higgses get the small ones.  It's not quite clear to me how that works.

Apparently a well-studied inflationary model.  I need more coffee, I'm not getting this.  But model has problem with either not lasting long enough, or generating too small a ns.  Trick to the solution: end inflation by falling into the wrong vacuum.  The thermal energy keeps us in that vacuum till the temperature drops below the vacuum energy, at which point we have a second inflationary period tunnelling to the true vacuum.

This also avoids cosmic strings associated with the breaking of the U(1); since in the false vacuum the U(1) is unbroken, they only form at the lower scale of breaking, and their production is related to the reheating scale.  Gravitino production benefits (is reduced) for the same reasons.

Tim likes AMSB, with a heavy (150 GeV) gravitino for the inflation _and_ for the Higgs mass.  That would seem to lead to a superpartner spectrum too heavy for the LHC …

Question: does the model have distinct predictions?  Answer: Cosmic strings at detectable levels.

3:30 pm: Philip Bechtie, "The Status of Constrained SUSY, and implications from the Higgs"

Why is the CMSSM still interesting?  "Only revert to more complex models if required" … I think you need to argue that it isn't, more strongly than you have.

Software: Fittino, publicly available C++ code covering several SUSY models and a number of measurements.  Seems to assume that we fit all DM with SUSY, perhaps over-constraining in that sense.

Simple theorist's tools seem to be in decent agreement with full LHC simulation limits.  That motivates the software's utility.

HiggsSignals designed to combine the various Higgs limits and measurements, in Fortran but at least the 2003 flavour.

This talk so far has had little to do  with SUSY.  It's been a good advert for this software tool instead.

Just as I say that, the CMSSM shows up.  Putting a chi-squared cut of less than 1000!  Please tell me there's more info coming, that's absurd.

Oh look, the minimum is just above the excluded region.  Which don't seem to be directly included in the plots.  That preference is … not sure why, actually.  g-2, perhaps.  Yes, seems to be the case.  Optimal CMSSM is now SM plus DM, hardly a surprise.

Still possible to have light sleptons and gauginos in this model.  Claim of chi2 per dof less than 1; call me skeptical.  Of course, I'm sure the Bayesian analysis would disfavour these models.

No comments:

Post a Comment